A conversation about scientific evolution

I recently had an interesting exchange with an anonymous poster called “Himself” on the talk.origins USENet group. I thought I’d put it up here with some links and the typos corrected. Since it’s visible on Google Groups, I’m not breaking any copyright here.

Himself:

Do there exist any real or meaningful similarities between natural selection and the processes by which scientists select among alternative hypotheses?

JSW:

Often asserted. The first person to draw the parallel was Thomas Henry Huxley; more recently Popper claimed just that (“our hypotheses die in our stead”). The best version of it is David Hull’s Science as a Process (1988), which is in dire need of a precis. I did my masters on it.

Himself:

Thanks for the ref. I just checked amazon.com. They don’t have a Kindle edition, so I’m going to read the online sample content and then will probably order the book.

JSW:

Can I suggest you find a copy in a library and copy the following chapters: 1, 2, 3 (which is historically wrong, but that doesn’t matter), 8 (which is the crucial chapter for your purposes), 9, 10, 11, 12 (also crucial), 13 and the conclusion. You can ignore the systematists war (which, since I know most of the players, I am assured is wrong in parts, although the systematists cannot agree on which parts) material. It’s basically a case study. Hull writes clearly, well and inspiringly. I have a paper revisiting some of his ideas here, published in Biology and Philosophy.

Himself:

Thanks again. Your paper answered some of my questions. It raised a few questions, too. Does the idea of theories (or lines of research?) as spatio-temporally individuated entities predict the existence of limits to the integration or consolidation of scientific ideas? Must these entities remain separate and distinct like individual biological species? Also, does Hull consider scientific research to be evolutionarily distinct from other sorts of cultural change?

JSW:

In order: No, because there are no set levels or ontologies of conceptual evolution. No, but when you have lineages that tend to travel together (like a set of political policies that are not in themselves integrated but which tend to be a particular party’s platform) you form what Toulmin referred to as traditions. These are the analogue of species in culture. It took me a masters and a PhD to resolve that question – what is the analogue of species in cultural evolution. And finally, no he doesn’t. Nor do I.

Himself:

I guess that science does impose a unique set of constraints on the types of ideas that can propagate within a population of interactors but, Insofar as nonscientists can adopt and modify ideas to suit their own purposes, the concept of conceptual inclusive fitness would seem applicable to other forms of cultural change.

JSW:

A science cannot get more complicated than it can be taught. Imagine a science so complicated it took forty years to learn. You’d have five productive years. No career could be built on that. So sciences tend to specialise into disciplines, subdisciplines, research programs and projects. They equilibrate around the social norm (which used to be about 14 years of education but is now around 18).

Himself:

OK. Thanks yet again. One more question occurs to me. The answer may be in your paper, but I haven’t yet read it thoroughly. Do you or Hull propose a relationship between the “meme-fitness” of theories and the biological fitness of scientists? And, if such a relationship does exist in any regular and discernable form, does it represent a special case of a broader relationship between innovation and biological reproductive success?

JSW:

We both explicitly reject that link. Hull has a discussion somewhere in SaaP about “Lamarckism” – that memes are not biological and so cultural evolution is not “Lamarckian” in the sense of inheritance of acquired characteristics. Memes, whatever they may be, are independent evolvers from genes and organisms. The tie-in I think has to do with the professional career of the individual – qua scientist, one’s career is good or ill; this has a link to how well the individual organism does in terms of resource acquisition; but memes can run counter to biological fitness (as in cases where one feels one has a “calling” to a career that actively depreciates one’s fitness; such as being the last member of a family and who decides to become a celibate).

Himself:

I was thinking in terms of someone, like Genghis Khan, who came up with new social and military innovations that allowed him to leave many highly successful descendants. I thought that scientific innovation might be based partly on genes that contribute to various forms of creativity or technical prowess, or even strong ambition.

JSW:

I’m not saying that it can’t; just that it rarely does, and the two domains are, in Toulmin’s term, “decoupled”.

Himself:

It’s that “decoupling” that bothers me. We have two domains in which remarkably similar selective processes seem to be operating. Yet, apart from the detail that memes inhabit our brains, the two domains are “decoupled.” So, what’s going on? Is “science as a selective process” merely a hugely overextended set of interrelated metaphors, or does this idea point to intrinsic aspects of science as a process? If it’s just a metaphor, it can be stretched a long way without becoming clearly ridiculous. In your paper I read of “demes” of scientists competing in what seems to be a “conceptual landscape” analogous to the fitness landscapes of biological evolution. I can’t tell at this point whether this is really a useful way of looking at science or merely a post-hoc rationalization imposed by you philosopher dudes.

JSW:

That’s the sociobiological/evolutionary psychological response: what we do in culture is constrained and possibly determined by our biology, and hence has a direct feedback to the biology. In my view this is an empirical, not a priori, question. As it happens I think that culture happens largely in the “slop” allowed by biology – just as we can individually adapt through learning and development to specific environments, without affecting in the short to medium term our genetic/phylogenetic adaptation, so culture happens in the short run. But that doesn’t mean that if the culture (or individual learning, for that matter) persists long enough we will not see population-level adaptation: cases like lactose tolerance, genes for disease resistance, and so on have spread through populations just in recent years (last 10ky), affecting the biology.

There is also the Baldwin Effect, in which adaptive peaks within individual adaptive reach (via learning and cultural transmission) can become assimilated by natural selection so that organisms do not have to learn. However, the rate of cultural persistence and novelty is, genetically speaking, too great to be either particularly constrained by or to affect biological evolution all that much; it is effectively just stochastic drift from a gene’s eye view. Migration has a greater effect than adaptation. So what scientists do is biologically no different to what religionists do, or artists, or day traders: it is a means of getting resources, which improves your individual and inclusive fitness, but it has its own evolutionary dynamic.

If this bothers you, think of it like this: evolution is like a sea with laminar flow layers: the deep layers move slowly, and the surface layers move rapidly, and they do interact at their layers of contact. However, they are their own flows at their own rates. A strong enough superficial layer can affect the deeper layers, yes, but only if it persists long enough.

Himself:

That seems a good metaphor. Our deep biological characteristics may have been largely determined hundreds of thousands of years ago. We don’t have to learn the deeper behaviors that influence – but don’t determine – our more superficial, and more rapidly changing, culturally influenced behaviors. It’s interesting that you mention religionists, artists and day traders together. Does it strike you that all have game-like aspects? I think I’ve heard a day trader explicitly call it his “game.” Is it possible that the tendency to form specialized game-like behaviors, within particular cultural contexts, is a species-specific Homo sapiens characteristic – one of those “deep layers” you mentioned?

JSW:

“Game”, famously regarded by Wittgenstein as a cluster of concepts and definitions, is a very good term for what humans do culturally, largely because it seems to me that we learn through game play, and culture is acquired through learning. Apprenticeships, such as scientists go through, are largely about learning the rules, the standards of success, and who is who in the process; this is a good specification of games. Humans acquire much of their culture as juveniles through game play; it is only extension of that play behaviour that much of the rest of culture is also a variety of game play too.